Children with Specific Language Impairment (SLI) are slower to learn new words than their peers, placing them at risk for academic failure. Our long-term goal is to develop an effective word learning treatment for kindergarten children with SLI, thereby improving their academic and vocational outcomes. During the prior funding period, we successfully taught new words to children with SLI via interactive book reading, a treatment involving an adult reading a storybook to a child and deviating from the text to teach new words. We identified the appropriate intensity of the treatment and showed that children with SLI can learn an appropriate number of words by the end of treatment. However, this successful support of short-term word learning revealed new challenges that must be overcome in this renewal to continue to understand and improve long-term word learning by children with SLI. Thus, a second preliminary clinical trial involving 60 kindergarten children with SLI is proposed in this renewal. Aim 1 addresses the challenge that newly learned words were forgotten once treatment was withdrawn. We attempt to buffer forgetting by comparing different amounts of testing during interactive book reading (low vs. mid vs high testing). Incorporating testing into training is a well-established and highly replicated means of reducing forgetting by adults and typically developing children. Aim 1 will determine whether testing can be harnessed to buffer forgetting by children with SLI under real world conditions. Aims 2 and 3 address the challenge that not all children benefitted equally from interactive book reading. In Aim 2, we identify pre-treatment characteristics of children with SLI that are associated with the slope of learning during treatment or the slope of forgetting post-treatment. Moreover, we select a pre- treatment battery that samples a wide array of skills likely to be associated with learning (language processing, working memory, and episodic memory) or forgetting (learning at last treatment, decay rate). Aim 2 will provide a foundation for predicting which children will benefit from interactive book reading and will identify which skills are major barriers to long-term word learning by children with SLI. In Aim 3, we classify each child?s response at the end of treatment (learner vs. non-learner) and at the end of post-treatment monitoring (rememberer vs. forgetter) and then examine earlier performance to determine when treatment and post-treatment outcomes can be predicted. This yields empirically based benchmarks for progress that can be used later to tailor the treatment to individual children and establishes the stability of learning and forgetting over time. Overall, this research advances a promising treatment to effectively overcome the significant word learning challenges faced by children with SLI and reveals the contribution of learning and forgetting to language normalization by children with SLI. The results will have impact beyond word learning and SLI because all treatments require boosting learning and buffering forgetting. Thus, the knowledge gained will further catalyze clinical research.